technique (6). Thus we concluded that the nonpermissive behavior of virus-infected mouse cells is not due to a lack of a cellular factor(s) required for late SV40 ...
JOURNAL OF VIROLOGY, Mar. 1976, p. 854-858 Copyright 0 1976 American Society for Microbiology
Vol. 17, No. 3 Printed in U.S.A.
Regulatory Mechanism of Simian Virus 40 Gene Expression in Permissive and in Nonpermissive Cells ADOLF GRAESSMANN,* MONIKA GRAESSMANN, AND CHRISTIAN MUELLER Institut fuer Molekularbiologie und Biochemie der Freien Universitaet Berlin, Berlin 33, Arnimallee 22, Germany Received for publication 12 September 1975
Primary or continuous lines of mouse cells (3T3) are nonpermissive for simian virus 40 (SV40). Abortively infected cells synthesize tumor antigen (T antigen) but not viral DNA and virus capsid protein (V antigen). V antigen, however, was obtained when SV40 DNA was injected into 3T3 cells. This late gene expression induced by SV40 DNA is gene dose dependent. Early SV40 gene expression also appears to be correlated with the quantity of injected DNA molecules per 3T3 cell. T antigen formation can be detected after microinjection of only 1 to 2 DNA molecules, but the intensity of intranuclear T antigen fluorescence is significantly brighter with injection of higher concentrations of viral DNA. In permissive cells (TC7), early and late SV40 gene expression is not related to the number of injected molecules. Microinjection of 1 DNA molecule induced T and V antigen formation with the same efficiency as microinjection of 2,000 to 4,000 molecules. The question of whether late SV40 gene expression is directly related to the quantity of an early virus-specific product was approached by microinjection of early SV40 complementary RNA together with small amounts of viral DNA. V antigen was obtained in a high proportion of recipient 3T3 cells at conditions where microinjection of viral DNA alone induced T but not V antigen synthesis. The infectious agent of simian virus 40 an early virus-specific product which can be obtained by microinjection of a high number of SV40 DNA molecules but not by infection of mouse cells using the conventional virus adsorption method. In the present study we have investigated this latter hypothesis by a direct experimental approach. We first determined the minimal number of SV40 DNA molecules required for T and V antigen formation in permissive (TC7) and nonpermissive (3T3) cells. In a subsequent series of experiments we microinjected 3T3 cells with amounts of SV40 DNA insufficient for induction of late viral gene expression together with in vitro synthesized early virus-specific mRNA (16; M. Graessmann and A. Graessmann, Proc. Natl. Acad. Sci. U.S.A., in press. V antigen formation was determined in these cells.
(SV40) is the viral DNA. The virus-coded information can be expressed in different manners, depending upon the cell type infected. Cells permissive for SV40 (monkey cells) support early and late viral gene expression, whereas nonpermissive cells (mouse cells) promote only expression of the early virus-specific functions (13). Early viral gene expression is indicated by the intranuclear tumor antigen (T antigen) formation and late gene expression by viral capsid protein (V antigen) synthesis (13). Recently we have shown that V antigen synthesis can also be obtained in primary mouse kidney cells when SV40 DNA component I (DNA I) is introduced by the microinjection technique (6). Thus we concluded that the nonpermissive behavior of virus-infected mouse cells is not due to a lack of a cellular factor(s) required for late SV40 gene expression but rather to restriction by a cellular repressing type factor(s) (14). The biological function of this proposed factor would be neutralized by microinjection of an excess of SV40 DNA molecules (6). However, these experiments did not exclude the possibility that late SV40 gene expression requires a threshold concentration of
MATERIALS AND METHODS Preparation of SV40 DNA I. Cultures of confluent TC7 cells, a subline of CV1 cells, grown in 100-mm plastic Petri dishes were infected with plaque-purified SV40 (777) at a multiplicity of 0.1 PFU per cell. After an adsorption period of 2 h, cells were washed and covered with serum-free Eagle medium. DNA was extracted from the cells by the 854
VOL. 17, 1976
EARLY SV40 mRNA AND LATE VIRAL GENE EXPRESSION
selective extraction method of Hirt (9). The extract was phenol treated (80% phenol in 1 M Tris-hydrochloride, pH 8.0), and nucleic acids present were precipitated with ethanol at -20 C. The precipitate was collected by centrifugation, washed with ethanol, dried, dissolved in O.1x SSC (1x SSC = 0.15 M NaCl plus 0.015 M sodium citrate), treated with RNase (Worthington, 50 Mg/ml, 30 min at 37 C), phenol treated, and reprecipitated with ethanol. The DNA was further purified by neutral sucrose gradient centrifugation (5 to 20% [wt/vol], 1 M NaCl, 0.001 M EDTA, 0.001 M Tris-hydrochloride, pH 7.4) and by CsCl-ethidium bromide equilibrium centrifugation (pi = 1.56 g of CsCl/ml, 100 Ag of ethidium bromide per ml in 0.02 M Tris-hydrochloride, pH 8.0) (7). Transcription of SV40 DNA I. Sigma factor containing DNA-dependent RNA polymerase was used for DNA transcription (Miles Laboratories, Escherichia coli K-12, RNA polymerase, grade II). The transcription reaction mixture (500 ul) contained 30 Mg of SV40 DNA I; 30 Mg of RNA polymerase; 0.15 M KCl; 0.04 M Tris-hydrochloride (pH 8.0); 5% glycerol; 0.01 M MgCl2; 1.5 mM 2-mercaptoethanol; ATP, GTP, CTP, and UTP, each at 0.4 mM; and 0.01 mM [5-'H ]UTP (Amersham Buchler, 15 Ci/mmol). The nucleoside triphosphates were added after preincubation of SV40 DNA and RNA polymerase at room temperature for 10 min. After 60 min of incubation at 37 C, 30 Mg of electrophoretically purified DNase (Worthington) was added to the reaction mixture, which was further incubated for 30 min at 37 C. Then 0.1x SSC and 5% sodium dodecyl sulfate were added to a final concentration of 0.01x SSC and 1% sodium dodecyl sulfate. The assay was extracted two times with 1 volume of phenolhydroxyquinoline (80% phenol in water [vol/vol], and 1% hydroxyquinoline [wt/vol]) and precipitated from the aqueous phase at -20 C with 2 volumes of ethanol (95%) and 1/10 volume of 1 M NaCl. To exclude any trace of SV40 DNA, the complementary RNA (cRNA) preparation was subsequently subjected to Cs2SO4 equilibrium density gradient centrifugation (2). The initial density in a solution of 1 mM EDTA0.01 M Tris (pH 7.2) was 1.62 g/ml. After centrifugation in a Spinco SW65 rotor at 30,000 rpm for 72 h at 5 C, the gradient was fractionated from the bottom of the tube, and fractions with a density between 1.62 and 1.65 g/ml were collected. Cs2SO4 was removed by passing the pooled fractions through a Sephadex G25 column and by exhaustive dialysis (2). The size of the cRNA was estimated by formaldehyde-sucrose gradient centrifugation (see Fig. 2) (10). Microinjection of nucleic acids. TC7 and 3T3 cells were grown in 100-mm plastic dishes on small glass slides (50 by 10 mm) subdivided in squares of 1 mm2. The transfer of nucleic acids was performed in injection buffer (phosphate buffer, pH 7.2) under a phase-contrast microscope at a 400-fold magnification by means of microglass capillaries having a diameter of 0.5 um at the tip. The volume injected per cell was10-1'to2 x 10- 1ml(3,4,8).
855
microinjected with SV40 DNA I. At a concentration of 1 mg of SV40 DNA per ml the average number of molecules transferred was about 2,000 to 4,000 per recipient cell (6). After microinjection cells were further cultivated in serum-free Eagle medium at 37 C, and V antigen formation was tested 24, 36 and 48 h postinjection by the indirect immunofluorescence
technique.
V antigen formation was first demonstrated 24 h after DNA injection. At this time 5 to 6% of the recipient cells exhibited a weak-positive intranuclear V antigen reaction. At 48 h after DNA injection 38% of the recipient cells showed a strong V antigen-specific fluorescence. On the contrary, 3T3 cells infected with plaque-purified wild-type virus (777) (500 PFU/cell) did not synthesize V antigen in a detectable amount, but about 90% of these cells exhibited T antigen formation. This abortive type of response is not altered in 3T3 cells preinjected with injection buffer alone before viral infection (500 PFU/ cell). The data shown in Table 1 indicate that V antigen synthesis is directly correlated with the number of SV40 DNA molecules injected per recipient cell. Microinjection of 1,000 to 2,000 DNA molecules induced V antigen formation in 25% and the transfer of 250 to 500 molecules in 4% of 3T3 cells. At a lower DNA concentration V antigen formation could not be demonstrated. To exclude the possibility that this gene dose effect on the formation of V antigen is due to an unspecific degradation of the injected DNA, polyoma virus (PV) DNA I was added to SV40 stock solutions, resulting in a final concentration of both SV40 and PV DNA of 1 mg per ml of injection buffer. As shown in Table 1, the ratio of recipient cells exhibiting V antigen synthesis corresponds to the quantity of SV40 DNA molecules but not to the absolute number of injected DNA molecules. Dose dependence of viral DNA for T and V antigen synthesis in TC7 and 3T3 cells. To obtain direct evidence of how many SV40 DNA molecules are necessary for early as well as for late SV40 gene expression in permissive (TC7) and nonpermissive (3T3) cells, viral DNA was injected at different concentrations into both types of cells, and T and V antigen formation was investigated. As summarized in Table 2, the formation of SV40 T and V antigen in TC7 cells did not depend upon the number of injected DNA molecules. Injection of 2,000 to 4,000 DNA molecules induced T and V antigen formation at the same ratio (100%) as injection of 1 to 2 DNA molecules. Moreover, the intenRESULTS sity of SV40-specific T and V antigen fluoresLate SV40 gene expression in 3T3 cells cence was similar at all DNA dilutions tested. microinjected with viral DNA. 3T3 cells were In 3T3 cells, however, the synthesis of V
856
GRAESSMANN, GRAESSMANN, AND MUELLER
J. VIROL.
TABLE 1. SV40 Vantigen formation in 3T3 cells microinjected with SV40 DNA or with a mixture of SV40 DNA and PV DNAa Injection of SV40 DNA I
Concn of SV40 DNA
Injection of a mixture of SV40 and PV DNA I
Concn of each DNA
(Ag/ml)
Avg no. of injected DNA molecules/cell
V antigen (
1,000 500 250 120 60
2,000-4,000 1,000-2,000 500-1,000 250-500 125-250
38 25 12 4 0
(l~g/ml)
1,000 500 250 120 60
0 500 750 870 940
Avg no. of injected DNA molecules/cell
V antigen
2,000-4,000 2,000-4,000 2,000-4,000 2,000-4,000 2,000-4,000
38 27 NDb 4 0
(M)
aEach point is based on a count of 300 injected cells and is reproducible within 10% of the given value. V antigen formation was investigated 48 h after DNA injection. "ND, Not done. TABLE 2. SV40 T and V antigen formation in TC7 and 3T3 cells microinjected with SV40 DNA Ia Concn of Avg no.DNA SV40 injected DNA
(DgNml)
molecules/cell
1,000 2,000-4,000 100 200-400 10 20-40 1 2-4 0.5 1-2 0.1 0.2-0.4
TC7 cells T an- V antigen tigen
100 100 100 100 100 20
100 100 100 100 100 20
3T3 cells T an- V an-
tigen tigen 100 100 100 100 80 4
38 4 0 0 0 0
a Each point is based on a count of 300 injected cells and is reproducible within ± 10% of the given value. The cells were fixed and stained for T and V antigen 48 h after injection.
antigen as well as the intensity of T antigen fluorescence was DNA dose dependent. V antigen synthesis required microinjection of at least 250 DNA molecules. T antigen formation was obtained by microinjection of 1 to 2 DNA molecules, but the intensity of the intranuclear T antigen-specific fluorescence was just detectable. At higher DNA concentrations the intensity of the T antigen fluorescence increased markedly and reached an intensity comparable to that of permissive cells after microinjection of 250 to 500 DNA molecules per 3T3 cell (Fig. 1). Early SV40 mRNA and late viral gene expression. Early SV40 mRNA was obtained by in vitro transcription of SV40 DNA I using DNA-dependent RNA polymerase from E. coli (16). After transcription the DNA template was removed from the reaction mixture by exhaustive DNase treatment and Cs2SO4 equilibrium density gradient centrifugation (2). The size of the cRNA was estimated by formaldehyde-
sucrose gradient centrifugation (10). Under the conditions used, the complete transcription product of one DNA strand would have a sedimentation coefficient of 26S (10). As shown in Fig. 2, about 50% of the in vitro synthesized RNA corresponds at least to the length of one SV40 DNA strand. Upon microinjection these cRNA preparations induced T antigen formation in 80% of the recipient 3T3 cells but no V antigen synthesis (Table 3). To show that early SV40 mRNA is involved in the process of late viral gene expression, 3T3 cells were microinjected with mixtures of SV40 DNA I and SV40 cRNA. At a constant concentration of 0.5 mg of SV40 cRNA per ml in the mixture, different concentrations of SV40 DNA I were chosen to transfer known amounts of DNA molecules into the cells. Injection of 20 to 40 DNA I molecules per cell yielded 50% V antigen-positive cells; when 2 to 4 DNA molecules were injected about 8% of the cells showed V antigen fluorescence (Table 3). These values were obtained after a 24-h incubation period, whereas incubation for 48 h resulted in a significant reduction of V antigen-positive cells (data not shown). Substitution of PV-specific cRNA in the injection mixture for SV40 cRNA did not affect SV40-specific T antigen formation, but V antigen synthesis was not detectable.
DISCUSSION
Although only one SV40 DNA I molecule is needed for highly efficient synthesis of early and late viral gene products in permissive TC7 cells, there exists a threshold concentration of about 250 to 500 DNA molecules for late viral gene expression in nonpermissive 3T3 cells (Tables 1 and 2). Moreover, our SV40 DNA dilution
VOL. 17, 1976
EARLY SV40 mRNA AND LATE VIRAL GENE EXPRESSION
857
FIG. 1. T antigen fluorescence in 3T3 cells microinjected with SV40 DNA I. (A) 2,000 to 4,000 DNA molecules per recipient cell; (b) 1 to 2 DNA molecules per recipient cell. Cells were fixed and stained 48 h after
injection.
experiments indicate that not only late but also early SV40 gene expression is correlated with the amount of viral DNA molecules transferred per cell as estimated by the intensity of the SV40-specific intranuclear T antigen fluorescence. These findings raised the question of whether late SV40 gene expression is directly related to the quantity of an early virus-specific product. Indeed, we were able to show that microinjection of early SV40 cRNA, which contains the information for T antigen and for stimulation of DNA synthesis (5; M. Graessmann and A. Graessmann, Proc. Natl. Acad. Sci. U.S.A., in press), together with 20 to 40 SV40 DNA molecules into 3T3 cells induced V antigen synthesis at a ratio not even obtained after injection of 2,000 to 4,000 DNA molecules alone. This effect is SV40 cRNA specific since PV cRNA cannot enter into this function (Table 2 and 3). There exists no direct experimental evidence
to indicate which of the early virus-specific functions is responsible for this regulatory mechanism. Since induction of viral DNA replication is a prerequisite for V antigen synthesis in permissive and nonpermissive cells and since early virus-specific mRNA contains the information for it, this function may be the candidate (1, 6). Evidence is accumulating that SV40 T antigen itself is the carrier of this regulatory function (11, 12, 15) and that the efficiency of its production may determine permissiveness and nonpermissiveness. Less efficient early viral gene expression in nonpermissive cells (3T3) could then be explained either in terms of positive or negative control or by the lack of an ability for processing of the early viral protein which may be a cellular requirement. Microinjection of this protein obtained either from infected cells or by translation of viral RNA in vitro should give a convincing answer.
GRAESSMANN, GRAESSMANN, AND MUELLER
858
J.- VIROL.
This work was supported by the Deutsche Forschungsgemeinschaft (Gr 384/2/4).
iS S28S 9
8 7
6 m
5
x E
CL 3
2
1
*
I
I
5
10
i
15
20
25
f raction no. FIG. 2. Velocity sedimentation of SV40 cRNA in sucrose density gradient (7 to 27% sucrose [wt/vol] in high-salt formaldehyde solution, 1.1 M HCHO, 0.09 M Na2HPO4, 0.001 M EDTA, pH 7.7). The gradient was centrifuged in a Spinco SW65 rotor at 49,000 rpm for 3 h at 20 C. Before centrifugation SV40 cRNA was heated for 5 min at 80 C in low-salt formaldehyde (1.1 M HCHO, 0.0045 M Na2HPO4, 0.0005 M NaHPO4. 0.001 M EDTA, pH 7.7).
TABLE 3. SV40 Tand V antigen formation in 3T3 cells microinjected with mixtures of SV40 DNA and cRNAa Concn of Avg no. V40 of injected DNA I DNA mole(ug/ml) cules/cell
10 10 1 0 10
20-40 20-40 2-4 0 20-40
Concn (ug/
SV40 T SV40 V
of cRNA antigen ml)(origin) antigen (% (% 0 500 (SV40) 500 (SV40) 500 (SV40) 500 (PV)
100 100 100 80 100
0 50 8 0 0
a Each point is based on a count of 300 injected cells and is reproducible within ± 10% of the given value. The cells were fixed and stained for T and V antigen 24 h after microinjection.
ACKNOWLEDGMENTS We are grateful to R. Bobrick, E. Guhl, and H. Koch for skillful technical assistance.
LITERATURE CITED 1. Cowan, K., P. Tegtmeyer, and D. D. Anthony. 1973. Relationship of replication and transcription of simian virus 40 DNA. Proc. Natl. Acad. Sci. U.S.A. 70:19271930. 2. Erikson, R. L. 1969. CsISO, banding of RNA, p. 460463. In K. Habel and N. P. Salzman (ed.), Fundamental techniques in virology. Academic Press Inc., New York. 3. Graessmann, A. 1970. Mikrochirurgische Zellkerntransplantation bei Saeugetierzellen. Exp. Cell Res. 60: 373-382. 4. Graessmann, A., and M. Graessmann. 1971. Ueber die Bildung von Melanin in Muskelzellen nach der direkten Uebertragung von RNA aus Harding-Passey-Melanomzellen. Hoppe-Seyler's Z. Physiol. Chem. 352:527-532. 5. Graessmann, A., M. Graessmann, E. Hoffmann, J. Niebel, G. Brandner, and N. Mueller. 1974. Inhibition by interferon of SV 40 tumor antigen formation in cells injected with SV 40 cRNA transcribed in vitro. FEBS Lett. 39:249-251. 6. Graessmann, M., and A. Graessmann. 1975. Regulation mechanism of simian virus 40 late gene expression in primary mouse kidney cells and simian virus 40 transformed 3T3 cells. Virology 65:591-594. 7. Graessmann, M., A. Graessmann, E. Hoffmann, J. Niebel, and K. Pilaski. 1973. The biological activity of different forms of polyoma virus DNA and viral DNA fragments. Mol. Biol. Rep. 1:233-241. 8. Gruen, R., M. Graessmann, A. Graessmann, and M. Fogel. 1974. Infection of human cells with polyoma virus. Virology 58:290-293. 9. Hirt, B. 1967. Selective extraction of polyoma DNA from infected mouse cell cultures. J. Mol. Biol. 26:365-369. 10. Mandel, J. L., C. Kedinger, F. Gissinger, P. Chambon, and A. H. Fried. 1973. Size of the RNA's synthesized by purified calf thymus DNA-dependent RNA polymerases SV 40 DNA. FEBS Lett. 29:109-112. 11. Reed, S. I., J. Ferguson, R. W. Davis, and G. R. Stark. 1975. T antigen binds to simian virus 40 DNA at the origin of replication. Proc. Natl. Acad. Sci. U.S.A. 72:1605-1609. 12. Tegtmeyer, P., M. Schwartz, J. K. Collins, and K. Rundell. 1975. Regulation of tumor antigen synthesis by simian virus 40 gene A. J. Virol. 16:168-178. 13. Tooze, J. (ed.). 1973. The molecular biology of tumor viruses. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 14. Watkins, J. F., and R. Dulbecco. 1967. Production of SV 40 virus in heterocaryons of transformed and and susceptible cells. Proc. Natl. Acad. Sci. U.S.A. 58:1396-1403. 15. Weil, R., C. Salomon, E. May, and P. May. 1975. A simplifying concept in tumor virology: virus-specific "pleiotropic effectors." Cold Spring Harbor Symp. Quant. Biol. 39:381-395. 16. Westphal, H. 1970. SV 40 DNA strand selection by Escherichia coli RNA polymerase. J. Mol. Biol. 50: 407-420.
